1,721,066 research outputs found
Justification in Case-Based Reasoning
Full text linkThe explanation and justification of decisions is an important subject in contemporary data-driven automated methods. Case-based argumentation has been proposed as the formal background for the explanation of data-driven automated decision making. In particular, a method was developed in recent work based on the theory of precedential constraint which reasons from a case base, given by the training data of the machine learning system, to produce a justification for the outcome of a focus case. An important role is played in this method by the notions of citability and compensation, and in the present work we develop these in more detail. Special attention is paid to the notion of compensation; we formally specify the notion and identify several of its desirable properties. These considerations reveal a refined formal perspective on the explanation method as an extension of the theory of precedential constraint with a formal notion of justification
Justifications derived from inconsistent case bases using authoritativeness
Full text linkPost hoc analyses are used to provide interpretable explanations for machine learning predictions made by an opaque model. We modify a top-level model (AF-CBA) that uses case-based argumentation as such a post hoc analysis. AF-CBA justifies model predictions on the basis of an argument graph constructed using precedents from a case base. The effectiveness of this approach is limited when faced with an inconsistent case base, which are frequently encountered in practice. Reducing an inconsistent case base to a consistent subset is possible but undesirable. By altering the approach’s definition of best precedent to include an additional criterion based on an expression of authoritativeness, we allow AF-CBA to handle inconsistent case bases. We experiment with four different expressions of authoritativeness using three different data sets in order to evaluate their effect on the explanations generated in terms of the average number of precedents and the number of inconsistent a fortiori forcing relations
Kettle logic in abstract argumentation
No full textKettle logic is a colloquial term that describes an agent’s advancement of inconsistent arguments in order to defeat a particular claim. Intuitively, a consistent subset of the advanced arguments should exist that is at least as successful at refuting the claim as the advancement of the set of inconsistent arguments. In this paper, we formalize this intuition and provide a formal analysis of kettle logic in abstract argumentation, a fundamental approach to computational argumentation, showing that all of the analysed abstract argumentation semantics (inference functions)—with the exception of naive semantics, which is considered a mere simplistic helper for the construction of other semantics—suffer from kettle logic. We also provide an approach to mitigating kettle logic under some circumstances. The key findings presented in this paper highlight that agents that apply the inference functions of abstract argumentation, are—similarly to humans—receptive to persuasion by agents who deliberately advance inconsistent and intuitively ‘illogical’ claims. As abstract argumentation can be considered one of the most basic models of computational argumentation, this raises the question to what extent and under what circumstances kettle logic-free argumentation can and should be enforced by computational means
Abstract Argumentation for Hybrid Intelligence Scenarios
No full textHybrid Intelligence (HI) is the combination of human and machine intelligence, expanding human intellect instead of replacing it. Information in HI scenarios is often inconsistent, e.g. due to shifting preferences, user's motivation or conflicts arising from merged data. As it provides an intuitive mechanism for reasoning with conflicting information, with natural explanations that are understandable to humans, our hypothesis is that Dung's Abstract Argumentation (AA) is a suitable formalism for such hybrid scenarios. This paper investigates the capabilities of Argumentation in representing and reasoning in the presence of inconsistency, and its potential for intuitive explainability to link between artificial and human actors. To this end, we conduct a survey among a number of research projects of the Hybrid Intelligence Centre. Within these projects we analyse the applicability of argumentation with respect to various inconsistency types stemming, for instance, from commonsense reasoning, decision making, and negotiation. The results show that 14 out of the 21 projects have to deal with inconsistent information. In half of those scenarios, the knowledge models come with natural preference relations over the information. We show that Argumentation is a suitable framework to model the specific knowledge in 10 out of 14 projects, thus indicating the potential of Abstract Argumentation for transparently dealing with inconsistencies in Hybrid Intelligence systems.Interactive Intelligenc
Empathic agents : A hybrid normative/consequentialistic approach
No full textComplex information systems operate with increasing degrees of autonomy. Consequently, such systems should not only optimize for simple metrics (like clicks and views) that reflect the system provider's preferences but also consider norms or rules, as well as the preferences of other agents that are affected by the systems' actions. As a means to achieve such behavior, we propose the design and development of empathic agents that use a mixed rule/utility-based approach when deciding on how to act, considering both their own and others' utility functions. The agents make use of formal argumentation to reach an agreement on how to act in case of inconsistent beliefs. A promising domain for applying our empathic agents is recommender systems.</p
Cautious nonmonotonicity
No full textA key intuition in symbolic artificial intelligence is that an intelligent system should be non-monotonic, but cautiously so: previous conclusions should only be revised if a compelling reason for doing so exists. In this paper, I trace the evolution of this intuition, which emerged from Dov Gabbay’s seminal 1985 paper and gained additional prominence as cautious monotonicity in the 1990 KLM paper, as well as in an earlier paper by Makinson. I introduce the term cautious nonmonotonicity for the general idea of assuring that monotonicity is satisfied given some condition, thus highlighting that it is the violation, and not the satisfaction, of monotonicity that we need to be careful about. Also, I discuss why cautious nonmonotonicity still is an open problem in theory and practice, and present some results that highlight the intricacy of cautious nonmonotonicity in the simple case of abstract argumentation, where inferences are drawn from directed graphs without further structure.Special Issue to Celebrate Dov Gabbay's 80th Birthday. ISBN: 978-1-84890-492-7</p
Cautious nonmonotonicity
No full textA key intuition in symbolic artificial intelligence is that an intelligent system should be non-monotonic, but cautiously so: previous conclusions should only be revised if a compelling reason for doing so exists. In this paper, I trace the evolution of this intuition, which emerged from Dov Gabbay’s seminal 1985 paper and gained additional prominence as cautious monotonicity in the 1990 KLM paper, as well as in an earlier paper by Makinson. I introduce the term cautious nonmonotonicity for the general idea of assuring that monotonicity is satisfied given some condition, thus highlighting that it is the violation, and not the satisfaction, of monotonicity that we need to be careful about. Also, I discuss why cautious nonmonotonicity still is an open problem in theory and practice, and present some results that highlight the intricacy of cautious nonmonotonicity in the simple case of abstract argumentation, where inferences are drawn from directed graphs without further structure.Special Issue to Celebrate Dov Gabbay's 80th Birthday. ISBN: 978-1-84890-492-7</p
Principbaserat icke-monotoniskt resonemang - från människor till maskiner
No full textA key challenge when developing intelligent agents is to instill behavior into computing systems that can be considered as intelligent from a common-sense perspective. Such behavior requires agents to diverge from typical decision-making algorithms that strive to maximize simple and often one-dimensional metrics. A striking parallel to this research problemcan be found in the design of formal models of human decision-making in microeconomic theory. Traditionally, mathematical models of human decision-making also reflect the ambition to maximize expected utility or a preference function, which economists refer to as the rational man paradigm. However, evidence suggests that these models are flawed, not only because human decision-making is subject to systematic fallacies, but also because the models depend on assumptions that do not hold in reality. Consequently, the research domain of formally modeling bounded rationality emerged, which attempts to account for these shortcomings by systematically relaxing the mathematical constraints of the formal model of economic rationality. Similarly, in the field of symbolic reasoning, approaches have emerged to systematically relax the notion of monotony of entailment, which stipulates (colloquially speaking) that when inferring a set of statements from a knowledge base, the addition of new knowledge to the knowledge base must not lead to the rejection of any of the previously inferred statements. By drawing from these developments in microeconomic theory and symbolic reasoning, this thesis explores different principle-based approaches to decision-making and non-monotonic reasoning. Thereby, abstract argumentation is used as a fundamental method for reasoning in face of conflicting knowledge (or: beliefs) that reduces non-monotonic reasoning to the problem of drawing conclusions (extensions) from a directed graph, and hence provides a neat abstraction for theoretical exploration. In particular, the works collected in this thesis i) introduce the consistent preferences property of microeconomic theory, as well as some relaxed forms of monotony of entailment as mathematical principles to abstract argumentation-based inference; ii) show how to enforce some of these principles in dynamic environments; iii) devise a formal approach to maximize monotony of entailment, given the constraints imposed by an inference function; iv) extend and apply the aforementioned approaches to the domains of machine reasoning explainability and legal reasoning.Digital ISBN missing in publication. </p
Economic rationality and abstract argumentation [Elektronisk resurs]
No full textThis article presents a line of work that builds a bridge between abstract argumentation as a method of non-monotonic reasoning and formal models of economically rational decision-making. As the foundation of this bridge, we introduce the reference independence principle, which is a key property of economic rationality, to abstract argumentation. We relate this principle to principles of non-monotonic reasoning and, from this starting point, outline a set of research directions we are pursuing to better integrate abstract argumentation and models of economic rationality.</p
Consistency Principles for Sequential Abstract Argumentation
No full textThis paper presents a principle-based perspective on ensuring consistency in sequential abstract argumentation, in which an abstract argumentation framework is iteratively resolved by determining its extensions and then (normally) expanded by adding new arguments and attacks (without changing attacks between existing arguments). As a starting point, we take reference independence – a key property in microeconomic theory – and derive an abstract argumentation principle from it to (roughly) stipulate that “if no new argument is accepted our conclusion remains the same”. Moreover, we introduce the cautious monotony principle, which can be colloquialized as “if no new argument at- tacks our conclusion we do not reject any part of this conclusion”. Cautious monotony is satisfied by some admissible set-based semantics and is not well-aligned with naive set-based seman- tics, whereas reference independence is satisfied by at least one of the naive set-based semantics (CF2 semantics), but is not well-aligned with the notion of admissibility.
- …
